US10575277B2 - Methods and apparatus for wideband localization - Google Patents
Methods and apparatus for wideband localization Download PDFInfo
- Publication number
- US10575277B2 US10575277B2 US16/162,085 US201816162085A US10575277B2 US 10575277 B2 US10575277 B2 US 10575277B2 US 201816162085 A US201816162085 A US 201816162085A US 10575277 B2 US10575277 B2 US 10575277B2
- Authority
- US
- United States
- Prior art keywords
- measurements
- signal
- backscatter node
- largeband
- reflective state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W64/00—Locating users or terminals or network equipment for network management purposes, e.g. mobility management
- H04W64/006—Locating users or terminals or network equipment for network management purposes, e.g. mobility management with additional information processing, e.g. for direction or speed determination
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
- G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
- G01S5/0205—Details
- G01S5/0221—Receivers
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
- G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
- G01S5/0205—Details
- G01S5/0221—Receivers
- G01S5/02213—Receivers arranged in a network for determining the position of a transmitter
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
- G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
- G01S5/0278—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves involving statistical or probabilistic considerations
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/26—Systems using multi-frequency codes
- H04L27/2601—Multicarrier modulation systems
- H04L27/2647—Arrangements specific to the receiver only
- H04L27/2649—Demodulators
- H04L27/265—Fourier transform demodulators, e.g. fast Fourier transform [FFT] or discrete Fourier transform [DFT] demodulators
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L5/00—Arrangements affording multiple use of the transmission path
- H04L5/0001—Arrangements for dividing the transmission path
- H04L5/0003—Two-dimensional division
- H04L5/0005—Time-frequency
- H04L5/0007—Time-frequency the frequencies being orthogonal, e.g. OFDM(A), DMT
Definitions
- a localization system detects 1D, 2D, or 3D position of a backscatter node or other object with high spatial accuracy and low latency.
- the LS may be employed to detect the 3D position of a moving object, and thus may facilitate fine-grained robotic tasks.
- the LS may localize backscatter nodes with sub-centimeter accuracy without constraints on their locations or mobility.
- the LS may extract a sensing bandwidth from every single backscatter packet that is three orders of magnitude larger than the backscatter communication bandwidth.
- the LS employs a Bayesian space-time super-resolution algorithm that combines time series of the sensed bandwidth across multiple antennas to enable accurate positioning.
- the LS simultaneously achieves a median accuracy of sub-centimeter in each of the x/y/z dimensions and a 99th percentile latency of less than 7.5 milliseconds in 3D localization.
- the LS achieved fine-grained positioning for agile robotic tasks, including in use scenarios with robotic arms and nanodrones and for indoor tracking, packaging, assembly, and handover.
- an LS performs RF-based identification and 3D localization for fine-grained robotic tasks.
- the LS achieves at least three technical goals: (i) spatial accuracy, (ii) ability to detect moving RFIDs, and (iii) scalability.
- the LS achieves sub-centimeter localization accuracy.
- This localization accuracy may facilitate agile manipulation tasks, such as grasping and packaging. For instance, this localization accuracy may enable a robot to align its grip with an object for item grasping and manipulation tasks.
- an LS tracks position and trajectory of a moving object with very low latency.
- the LS may be fast enough to track mobile objects that are moving in random mobility patterns, and thus may facilitate spatially precise manipulation of objects by a robotic arm.
- an LS system may employ inexpensive hardware (such as inexpensive RFID tags).
- inexpensive hardware such as inexpensive RFID tags.
- the LS may cost-effectively scale to hundreds or thousands of items, even in relatively confined areas like homes or small businesses.
- WiFi®-based and Bluetooth®-based solutions are not scalable in our context since it is not feasible or cost-effective to tag every item with a Bluetooth® or WiFi Radio®.
- billions of manufactured items are already tagged with few-cent RFIDs, making RFID-based localization attractive from a scalability standpoint.
- conventional RFID localization solutions may be limited either in their accuracy or in their ability to deal with mobility.
- Conventional RFID techniques such as Tagoram and RFIDraw may work with mobile RFIDs, but they have decimeter-scale accuracy in obtaining a tag's exact position.
- an LS performs RF localization that achieves all three of the above goals.
- the LS has at least two technical innovations that together allow it to achieve high accuracy and random mobility
- the LS estimates an RFID's response over a wide bandwidth—e.g., one that is at least three orders of magnitudes larger than the backscatter communication bandwidth.
- the large bandwidth may be employed to mitigate reflections from other objects in the environment and to isolate the RFID's response.
- the LS may estimate the wide bandwidth in one shot from every single RFID response.
- the LS may include a wideband localization helper which acts like a radar. The helper may transmit a wideband signal, and may measure its reflection off different objects in the environment, including an RFID tag.
- the LS may transmit the helper's wideband signal in such a way that the wideband signal is compatible with the RFID tag's protocol, and may synchronize the helper with an RFID reader to isolate the RFID's reflection and estimate its wideband channel from every single response.
- the one-shot estimation enables accurate and precise positioning at speeds where conventional systems fail.
- the LS employs a wideband of sensing frequency—e.g., a sensing bandwidth that is greater than 10 MHz, or greater than 50 MHz, or greater than 100 MHz, or greater than 200 MHz.
- Bayesian Space-Time Super-Resolution Unfortunately, with conventional localization methods, the bandwidth that may be estimated from an RFID tag remains limited by the RFID tag's antenna response and impedance matching circuitry. As a result, the ability to sense an RFID's channel with conventional localization methods significantly degrades beyond a couple hundred MHz, even if the conventional method performs frequency hopping. Such bandwidth is still an order of magnitude lower than that required for sub-centimeter localization using ultra-wideband techniques.
- an LS performs a space-time super-resolution algorithm that overcomes this challenge.
- This localization algorithm may employ a wide bandwidth of frequencies (e.g., more than 100 MHz) to narrow down the potential locations of an RFID tag to a handful of candidates. By combining these candidates over space (multiple antennas) and time, the LS may zoom in on the exact location.
- the localization algorithm may solve a Gaussian mixture problem and incorporate it into a Bayesian framework that fuses spatio-temporal bandwidth measurements. Further, to deal with the nonlinear nature of the measurements, the LS may perform an approximate inference algorithm that exploits RF and geometric properties of the underlying estimators to achieve a computationally efficient solution for highly accurate positioning.
- the LS includes USRP (universal software radio peripheral) X310 software radios and detects the position and trajectory of off-the-shelf, battery-free RFID tags.
- USRP universal software radio peripheral
- X310 software radios detects the position and trajectory of off-the-shelf, battery-free RFID tags.
- the LS's 99th percentile accuracy remains lower than 2 cm in each of the x/y/z dimensions, while retaining a 99th latency smaller than 7.5 milliseconds in 3D localization.
- the LS (a) accurately tracked collaborative tasks between robotic arms including packaging, handover, and assembly; and (b) accurately tracked nanodrones as they flew in indoor environments.
- an LS performs one-shot wideband estimation from every backscatter packet, enabling low-latency and high-precision localization.
- the LS may perform a spatio-temporal Bayesian framework for RF localization that fuses time series of bandwidth and phase measurements across multiple antennas.
- the LS may perform an approximate inference algorithm with fast convergence time for RF localization.
- the algorithm may achieve computational efficiency by incorporating the properties of RF signals and the geometric nature of its measurements.
- an LS tracks position and trajectory of objects while the objects are manipulated by robotic arms, and thus facilitates highly precise robotic manipulation.
- an LS performs full-body motion tracking using battery-free RFID tags for virtual and augmented reality applications. For instance, multiple RFID tags that are positioned on different limbs on a user may be tracked, during full-body tracking.
- the LS performs full-body tracking for rehabilitation, sports analytics, or gaming.
- the LS may perform an algorithm that take as inputs measurements over a large bandwidth of reflected RF radiation from a backscatter node, enabling sub-centimeter localization accuracy.
- the LS tracks multiple points on the user's body with sub-20-millisecond latency, which is sufficiently fast for VR/AR (virtual reality/augmented reality).
- FIG. 1A shows a bottle and bottle cap, each with an RFID tag attached.
- FIG. 1B shows robotic arms screwing the bottle cap unto the bottle.
- FIG. 1C is a chart that shows plots of estimated position and actual position of RFID tags.
- FIG. 2 shows hardware for precisely determining the position of one or more RFID tags.
- FIG. 3 illustrates a localization method
- FIG. 4 illustrates one-shot localization with orthogonal frequency-division multiplexing (OFDM).
- FIG. 5 illustrates selecting OFDM symbols that occur entirely during a period in which a backscatter node is not transitioning between reflective states.
- FIG. 6 shows a set of ellipses that are indicative of possible locations of a backscatter node.
- FIG. 7 shows a Hidden Markov Model
- FIG. 8 illustrates centroid approximation
- FIG. 9 shows normals at an intersection of ellipses.
- FIGS. 10A and 10B illustrate a particle filter converging on an accurate estimate of location.
- FIGS. 11A and 11B together comprise a flowchart of a localization method.
- an LS performs ultra-low latency, very high accuracy localization of backscatter sensors. This localization, in turn, may facilitate fine-grained robotic applications. In illustrative implementations, the LS accurately performs localization both in line-of-sight and through occlusions.
- the LS operates with inexpensive backscatter sensors such as off-the-shelf RFID tags without hardware modifications to the tags; and (b) is fully compatible with a standard UHF RFID protocol.
- the backscatter sensor e.g., RFID tags
- the backscatter sensor may be attached as stickers to objects of interest (e.g., manufacturing items) for object manipulation or may be affixed to miniature robots like nanodrones for tracking.
- FIGS. 1A, 1B, and 1C taken together, show an illustrative use case of this invention, in which the position of RFID tags is determined with great precision and with very low latency.
- the objects being robotically manipulated are a bottle and a bottle cap.
- FIG. 1A shows a bottle 105 and a bottle cap 107 .
- RFID tag 101 is affixed to bottle 105 and RFID tag 103 is affixed to cap 107 .
- FIG. 1B shows robotic arms 114 , 116 screwing the bottle cap 107 unto the bottle 105 .
- the robotic arms include pincers 110 , 112 that grasp the bottle cap 107 bottle 105 .
- the robotic manipulation shown in FIG. 1B may be enabled by precise, low latency localization of the RFID tags 101 , 103 in accordance with the present invention.
- FIG. 1C is a chart that shows plots of estimated position and actual position of RFID tags.
- FIG. 1C plots: (a) actual trajectory 130 of RFID tag 101 on bottle 105 ; (b) estimated trajectory 131 of RFID tag 101 on bottle 105 ; (c) actual trajectory 132 of RFID tag 103 on bottle cap 107 ; and (d) estimated trajectory 133 of RFID tag 103 on bottle cap 107 .
- the estimated trajectories were estimated by a prototype of this invention. As may be seen from FIG. 1C , this invention may estimate position with sub-centimeter precision.
- FIG. 2 shows hardware 200 for precisely determining the position of one or more RFID tags, in an illustrative implementation of this invention.
- a transmit (Tx) antenna of an RFID reader 201 transmits a query 230 to a RFID tag 210 .
- Tag 210 reflects a modulated response 231 to a receiver (Rx) antenna of RFID reader 201 .
- a wideband transceiver (sometimes referred to herein as a “localization helper”) 202 transmits, via one or more transmit antennas, OFDM symbols 232 to RFID tag 210 .
- each OFDM symbol includes multiple orthogonal subcarriers in different frequency bins, where the highest frequency of the highest frequency subcarrier is much greater than the lowest frequency of the lowest frequency subcarrier. For instance, the difference between these highest and lowest frequencies may be greater than or equal to 10 MHz, or greater than or equal to 50 MHz, or greater than or equal to 100 MHz, or greater than or equal to 200 MHz.
- a controller computer 203 controls and interfaces with the RFID reader 201 and wideband transceiver 202 .
- controller 203 may comprise a microcontroller.
- an additional computer 220 may control and interface with controller 203 .
- Computer 220 may store data in, and access data stored in, a memory device (e.g., 221 ).
- controller 203 may store data in, and access data stored in, a memory device.
- a human user may input data or instructions to computer 220 —and may receive outputted data from computer 220 —via one or more I/O (input/output) devices 222 .
- the one or more I/O devices 222 may comprise any combination of one or more display screens, touch screens, keyboards, computer mice, microphones, speakers, earbuds, earphones, projectors, and haptic transducers.
- controller 203 or computer 220 controls transmission of RF signals by reader 201 and by wideband transceiver 202 (or performs post-processing of measurements of received RF signals) in such a way that location of RFID tag 210 is estimated based on measurements that are taken while RFID tag 210 is in a first reflective state or is in a second reflective state and that are not taken while RFID tag 210 is transitioning between the first and second reflective states.
- RFID tag 210 may be more reflective of RF radiation during the first reflective state than during the second reflective state, or may reflect RF radiation at a different phase during the first reflective state than during the second reflective state.
- changes in the reflective state of RFID tag 210 are caused by changes in impedance of the RFID tag 210 .
- the RFID tag may switch rapidly between: (a) a more reflective state and (b) a less reflective state.
- a switch in the RFID tag may be closed, causing the tag's antenna to be connected to ground, impedance in a circuit that includes the tag's antenna to be zero (or close to zero), and more RF power to be reflected by the tag's antenna.
- the switch in the RFID tag may be open, causing RF power to flow into the tag's power harvesting unit, impedance in a circuit that includes the tag's antenna to be high, and less RF power to be reflected by the tag's antenna.
- the switch in the RFID tag that is employed for this modulation may comprise a transistor.
- the reflective state of a backscatter node may vary over time, in such a way that the backscatter node is, at any given time, in either a first reflective state or a second reflective state or is transitioning between the first and second states.
- the first reflective state may differ from the second reflective state in that a reflection response of the backscatter node is different in the first reflective state than in the second reflective state.
- the reflection response may be phase or amplitude of RF reflection from the backscatter node as a function of RF radiation incident on the backscatter node
- the impedance (and thus reflective state) of RFID tag 210 is modulated by RFID tag 210 in response to an RFID signal transmitted by RFID reader 201 .
- reflective state of RFID tag 210 is being modulated (in response to an RFID signal transmitted by RFID reader 201 ): (a) one or more transmit antennas of wideband transceiver 202 may transmit the OFDM symbols; and (b) modulated OFDM symbols (e.g., h 2 (t) 233 ) may reflect from RFID tag 210 and travel to one or more receive antennas of the wideband transceiver 202 (sometimes called the “localization helper” herein).
- the RFID query 230 and RFID response 231 are in a narrow ISM frequency band (e.g.
- the OFDM symbols are transmitted in a much wider frequency band (e.g., that has a bandwidth of at least 50 MHz or at least 100 MHz or at least 200 MHz.)
- each specific receiver antenna of the wideband transceiver 202 receives both (a) a modulated signal (e.g., h 2 (t) 233 ) that reflects from RFID tag 210 and (b) a signal h 1 (t) that is travels directly from a transmit antenna of wideband transceiver 202 .
- a modulated signal e.g., h 2 (t) 233
- a signal h 1 (t) that is travels directly from a transmit antenna of wideband transceiver 202 .
- one or more of reader 201 , controller 203 or computer 220 may extract signal h 2 (t) 233 received at a particular receiver antenna of the wideband transceiver 202 from the total signal received at that antenna.
- an LS tracks location of multiple RFID tags (e.g., 210 , 211 , 212 ). To do so, RFID reader 201 may perform RFID singulation to detect the identity of each RFID tag based on the tag's uniquely modulated response. The localization of the different RFID tags may be performed in rapid sequence, one RFID tag at a time.
- the system hardware 200 includes an RFID reader 201 and a wideband transceiver (sometimes called a “localization helper” herein) 202 .
- the reader and helper may be employed together to obtain wideband estimates of position of RFID tag 210 .
- the LS includes an RFID reader and localization helper (also called a wideband transceiver) as shown in FIG. 2 .
- the helper may transmit wideband signals, and may capture their reflections off different objects in the environment. Taken together, the RFID reader and wideband helper may enable both identification (through the RFID's identifier) and accurate localization (using the wideband signals).
- the LS may include a centralized controller (e.g., 203 in FIG. 2 ) that (a) synchronizes the helper's signals with the RFID reader at the physical layer; and (b) on the protocol level, incorporates the helper's operation into the reader's finite state machine.
- the LS leverages the wideband channel estimates for localization.
- the LS may fuse estimates across multiple receive antennas of the localization helper and across time through a spatio-temporal Bayesian framework.
- the LS may model each antenna's measurements as a Gaussian mixture, and may solve for each tag's location and trajectory through an approximate inference algorithm.
- the approximate inference algorithm may linearize and approximate the antenna measurements in 3D and may propagate their beliefs through a particle filter. In some use scenarios, the resulting accuracy is high enough to enable the LS to track fine-grained robotic tasks.
- FIG. 3 illustrates a localization method.
- the method includes: (a) fusing wideband estimates across space and time through a Bayesian framework; and (b) solving for accurate 3D positions of a backscatter node by using approximate inference.
- wideband estimates 302 may be derived from measurements taken by a wideband transceiver (sometimes called a “localization helper” herein) 301 .
- the wideband estimates are estimates of different subcarriers, each as received at different times or at different antennas or at both different times and different antennas.
- approximate inference 320 is employed to determine localization of a backscatter node.
- the approximate inference algorithm 320 (a) takes, as inputs, the wideband estimates (e.g., 303 , 304 , 305 ); (b) models probability distributions of possible distances, by employing mixtures of gaussians (“MoG”) (e.g., 306 , 307 , 308 ); (c) combines the MoGs into a 3D MoG 310 ; (d) solves a Hidden Markov Model by employing sequential importance sampling 311 , and (e) precisely estimates a 3D trajectory and 3D positions of a moving backscatter node 312 .
- MoG mixtures of gaussians
- FIG. 4 illustrates one-shot localization with orthogonal frequency-division multiplexing (OFDM).
- a transmitter 401 in a localization helper transmits OFDM symbols.
- a receiver 403 in the localization helper demodulates these symbols by performing an FFT (fast fourier transform) followed by an overall channel estimation step.
- the overall channel is fed into an RFID packet and edge detection module, in order to discover the RFID's state transitions. Based on the output of the edge detection and elimination block, reliable wideband channel estimates may be extracted.
- FFT fast fourier transform
- an LS obtains wideband estimates from every single (narrowband) RFID response.
- the one-shot estimation enables the LS to track changes in the channel at very high speeds.
- the LS may employ the wideband estimates for accurate localization.
- a wireless device In backscatter networking, a wireless device called a reader may start a communication session by sending a signal on the downlink.
- a backscatter sensor e.g., an RFID tag
- the sensor may switch between two states: reflective and non-reflective, to transmit bits of zeroes and ones.
- Backscatter modulation may be frequency agnostic: e.g., an RFID tag may modulate not only the reader's signal but also all transmitted signals in the environment. This, in turn, may enable the LS to estimate an RFID's channel out-of-band.
- a wideband transceiver sometimes called a “localization helper” herein
- a technical problem which the inventors solved is: How to estimate a wide bandwidth in a single shot, e.g., from every single RFID response. In principle, one could do that by simply transmitting a wideband signal from the localization helper simultaneously with the reader, and estimating the channel across its bandwidth.
- this approach is complicated by two technical problems: First, the RFID's switching process introduces a fast fading channel for the wideband signals; ignoring such fading corrupts the wideband estimates. Second, by spreading its transmitted signal over a wide bandwidth rather than a single frequency, the helper's signal-to-noise ratio (SNR) significantly degrades, which limits its ability to detect the RFID's response and precludes channel estimation.
- SNR signal-to-noise ratio
- the localization helper (of the LS) overcomes these two technical problems. Among other things, we describe how, in some implementations: (a) the helper's wideband signals are transmitted in such a way as to enable backscatter-aware channel estimation; and (b) the RFID tag's response is robustly detected at the wideband receiver and then employed to accurately estimate channels.
- the LS performs an OFDM (orthogonal frequency division multiplexing) modulation technique.
- the OFDM modulation may simplify channel estimation over a wide bandwidth.
- a wideband channel may be divided into a set of multiple narrowband channels, and OFDM modulation may be performed in the frequency domain.
- the OFDM modulator may encode information in the frequency domain as X(f) and then take an IFFT (inverse fast fourier transform) before transmitting the signal over the air;
- the receiver may demodulate the signal by taking a FFT (fast fourier transform), and may estimate the channel H(f) by dividing the FFT's output by X(f).
- the OFDM symbols are transmitted in such a way that the symbols may be employed with backscatter modulation
- FIG. 5 shows an example of both backscatter modulation and the OFDM symbols over time.
- the entire OFDM symbol lies within a channel coherence time (i.e., that the channel does not change during the estimation process).
- a backscatter node switches its impedance, it causes an extremely fast-fading channel and corrupts the OFDM channel estimate.
- we choose long OFDM symbols then all of them will be corrupted with the backscatter switching process.
- we choose very short symbols we cannot pack many frequencies into them, which would prevent us from estimating the wideband channel at sufficient frequency resolution to deal with frequency selectivity.
- the LS may exploit information from the RFID reader about the backscatter switching rate. Specifically, the reader may communicate that rate (called backscatter link frequency or BLF) in its downlink command to the backscatter node. Hence, by coordinating with the RFID reader, the localization helper may use the BLF to construct its OFDM symbols.
- BLF backscatter link frequency
- an OFDM symbol may be transmitted in such a way that it lies entirely within a specific RFID reflection mode (i.e., transition-free region).
- the OFDM symbol duration T symbol may be smaller than half the period of an RFID switching period T switching :
- T symbol ⁇ T switching 2 1 2 ⁇ ⁇ BLF
- an OFDM symbol consists of N samples (i.e., N subcarriers), each with time period T sample , and (b) B denotes the overall bandwidth of the helper's OFDM transmission.
- the LS may employ a number N of samples (OFDM subcarriers) such that
- the LS may employ OFDM where N is less than 100.
- the localization helper may obtain robust channel estimates on the receive side. Recall that a difficulty in wideband estimation arises from the low SNR at each subcarrier since the power is spread across frequencies. This may complicate packet detection and reduce the reliability of the sensed channel.
- the LS exploits a frequency agnostic property of backscatter modulation.
- the LS may incoherently average their estimates. Such averaging may enable the LS to reliably observe changes in the overall reflected power and to use them to detect the beginning of the RFID response. Specifically, for every OFDM symbol at time n, the LS may compute
- the LS may leverage knowledge of the RFID packet's preamble to detect the packet start. Specifically, every RFID packet payload may be preceded by a known preamble p(n). Hence, the localization helper may correlate the averaged channel estimates H combined (n) with the preamble to detect packet start. This correlation may be written as:
- the helper may identify the packet beginning when D rises above a threshold.
- the LS may eliminate corrupted OFDM symbols. Recall the LS may construct the OFDM symbols to accommodate for the backscatter reflection rate. While this may ensure that at least one whole symbol is in a reflective or non-reflective state, it does not ensure that all OFDM symbols are non-corrupted.
- the localization helper may automatically detect and discard erroneous channel estimates. In doing so, the helper may retain only the channel estimates that are obtained when the RFID is reflecting or not reflecting, while eliminating estimates corrupted by RFID state transitions.
- FIG. 5 illustrates selecting OFDM symbols that occur entirely during a period in which a backscatter node is not transitioning between reflective states.
- a modulated signal 500 reflects from a backscatter node.
- the backscatter node is in a first reflective state.
- the backscatter node is in a second reflective state.
- the backscatter node may be more reflective (and thus amplitude of modulated signal 500 is greater) during the first reflective state than during the second reflective state.
- the backscatter node may transition between the first and second reflective states.
- the OFDM symbols 520 are each shorter in duration than a single time interval (e.g. 501 , 503 , 505 ) throughout which the backscatter node remains in the first reflective state; and (b) are each shorter than a single time interval (e.g. 502 , 504 ) throughout which the backscatter node remains in the second reflective state.
- the duration of each OFDM symbol may be shorter than one half of the multiplicative inverse of the backscatter link frequency of the backscatter node.
- At least a subset of the OFDM symbols may occur entirely during a period in which a backscatter node is not transitioning between reflective states.
- OFDM symbols 521 and 522 may each occur entirely during period 501 while the backscatter node remains in the first reflective state.
- Edge transitions e.g., 511 , 512 , 513 , 514 , 515
- may occur during a subset of the OFDM symbols e.g., during OFDM symbols 523 , 524 , 525 , 526
- These corrupted channel estimates may be disregarded when computing position or trajectory of the backscatter node.
- the OFDM symbols may be transmitted only at times such that (a) each of the transmitted OFDM symbols occurs entirely in a first reflective state or in the second reflective state; and (b) an edge transition does not occur during any of the transmitted OFDM symbols.
- RFID tags may communicate by switching between reflective and non-reflective states.
- the localization helper constructs and decodes its OFDM symbols to accommodate the backscatter switching.
- the LS estimates the RFID channels at each of the subcarriers.
- the OFDM channel estimates H(f) comprise not only the RFID's reflection but also the direct path between the helper's transmit and receive antennas as well as other reflections in the environment.
- the LS may exploit the fact that the difference between the reflective and non-reflective states is due to the RFID, and may subtract them from each other to obtain the RFID's channel.
- the helper is left with L symbols where the RFID is non-reflective and M symbols where it is reflective.
- the channels may be estimated at each subcarrier f as:
- the average may be calculated based on the reflective states and non-reflective states of a single RFID response, enabling one-shot wideband estimation.
- the localization helper may be taken into account in the finite state machine (FSM) of the RFID reader. This, in turn, may enable the localization helper to perform the OFDM-based demodulation over only the short interval of time during which the RFID's response is expected, rather than over the entire duration of the reader's communication session.
- the receiver may open a short time window immediately after the reader finishes transmitting its query command. For instance, this window may be 300 ⁇ s-long in comparison to the 2 ms-long communication session.
- this synchronous architecture may save significant computational resources (e.g., by 6.6 ⁇ ), allowing the LS to achieve ultra-high frame rates and ultra-low latency.
- the helper's transmitter and receiver are connected to the same oscillator; and (b) thus the estimated channels do not have any carrier frequency offset (CFO) or sampling frequency offset (SFO).
- CFO carrier frequency offset
- SFO sampling frequency offset
- each OFDM symbol may act as a cyclic prefix for the subsequent one.
- the helper's transmit and receive antennas are co-located;
- the line-of-sight signal dominates the channel estimate, and
- sampling offset (or packet detection delay) between the transmitted and received OFDM symbols is minimal or non-existent.
- the LS may perform a time-domain correlation that detects the beginning (i.e., first sample) of the first OFDM symbol. Moreover, the LS may perform a one-time calibration with a known RFID location in order to eliminate other over-the-wire hardware channels.
- the localization helper may drop the DC subcarrier because the DC subcarrier is less robust to noise, and because dropping the DC subcarrier may improve the dynamic range of the ADC (Analog-to-Digital Converter).
- ADC Analog-to-Digital Converter
- the LS may employ an Exponential Moving Average (EMA) on the channel estimates.
- EMA Exponential Moving Average
- the EMA may provide more robust channel estimates without sacrificing mobility. This may enable the LS to achieve higher accuracy.
- UWB ultra-wideband
- conventional UWB systems leverage their large bandwidth to compute the time-of-flight—i.e., the time it takes their signals to travel between a transmitter and a receiver.
- Conventional UWB systems then map this time-of-flight to the distance traveled by multiplying it by the speed of propagation. Because time and frequency are inversely related, the distance resolution is inversely proportional to their bandwidth.
- backscattering 30 GHz off RFID tags is neither feasible nor desirable for multiple reasons.
- backscatter devices have antennas and impedance matching circuits that are optimized to harness energy within a specific frequency band. Hence, signals backscattered significantly outside their optimal frequency band have very poor SNR, making channel estimation infeasible.
- ⁇ 2 ⁇ ⁇ ⁇ d ⁇ ⁇ ⁇ mod ⁇ ⁇ 2 ⁇ ⁇ ⁇
- d the distance traveled
- ⁇ the wavelength
- a difficulty in using the phase is that it wraps around every wavelength. Hence, it allows us to accurately recover a fractional distance d frac modulo the wavelength. Said differently, we would know that the actual distance is d frac +n ⁇ , but n may be an unknown integer.
- the LS obtains a wide bandwidth and the phase from every antenna, and combines them to narrow down the candidate locations to a handful.
- the LS employs a Bayesian framework to fuse measurements across space and time:
- each distance candidate may map to an ellipse whose foci are the transmit and receive antennas. This is because the LS's antennas measure the round-trip distance from the transmitter to the backscatter tag, and back to the receiver. Since a given transmit-receive pair has multiple distance candidates, this leads to multiple confocal ellipses as shown in FIG. 6 . Adding more receive antennas would create other sets of ellipses. However, due to noise, we do not expect the correct ellipses from all antennas to intersect at the same point. Thus, in many cases, we do not simply rely on voting to identify the correct candidate.
- FIG. 6 illustrates a set of ellipses that are indicative of possible locations of a backscatter node.
- distances map to confocal ellipses.
- ellipse 611 consists of all points p such that the shortest RF pathlength from transmitter 601 to p and then to receiver 602 is equal to d
- ellipse 612 consists of all points q such that the shortest RF pathlength from transmitter 601 to q and then to receiver 602 is equal to d+
- ellipse 613 consists of all points r such that the shortest RF pathlength from transmitter 601 to r and then to receiver 602 is equal to d+2 ⁇ ; where d is a distance calculated from phase and A is wavelength.
- the different pathlengths are each determined based on measured phase, but differ from each other due to phase wrapping (i.e.,
- each of the candidate locations may trace a different trajectory.
- An advantage of using temporal series is that because noise is random, we expect the intersection points that correspond to the actual location to be closer to each other across time, providing opportunities to identify them.
- HMM Hidden Markov Model
- the localization problem is formulated as a Hidden Markov Model (HMM), as shown in FIG. 7 .
- HMMs form a class of powerful Bayesian inference models with hidden states and observed variables.
- the hidden states correspond to the actual locations of the RFID over time, and the observations are the candidate distances obtained from the different receive antennas.
- the LS's MINI has a nonlinear Gaussian observation model and a linear Gaussian transition model: the distance observations are nonlinear in the state variables (the coordinates) as may be seen with the function, while state transitions are linearly related due to motion.
- the nonlinearity may prevent us from adapting common solutions like Kalman Filters to model the distributions.
- FIG. 7 illustrates a Hidden Markov Model (HMM), in an illustrative implementation of this invention.
- the hidden states are positions of a backscatter node over time.
- hidden states 701 , 702 , 703 are positions x(t ⁇ 1), x(t), and x(t+1), respectively (that is, the position of the backscatter node at times t ⁇ 1, t, and t+1).
- observed states 704 , 705 , 706 are measurements y(t ⁇ 1), y(t), and y(t+1), respectively (that is, measurements at times t ⁇ 1, t, and t+1).
- the hidden states may be inferred from the observed states.
- transitions between hidden states 701 , 702 , 703 may be linear due to motion; and (b) transitions between observed states 704 , 705 , 706 may be non-linear.
- the observed states (measurements) may be modeled as gaussian related to each other over time.
- the LS's goal is to find the most likely trajectory x 1 . . . x T , given the observations (distance candidates) y 1 . . . y T .
- the LS may solve for the maximum a posteriori (MAP):
- x 0 ⁇ : ⁇ T * arg ⁇ ⁇ max x 0 ⁇ : ⁇ T ⁇ ⁇ p ⁇ ( x 0 ⁇ : ⁇ T ⁇
- x t is a d dimensional coordinate vector of the tag
- y t is a k dimensional vector of the estimated distances from each receiver
- y) denotes the posterior probability distribution.
- x i ) and a prior p (x 0:T ) may be modeled.
- the posterior may be expressed as: p(x 0:T
- the likelihood function may be modeled as follows: Recall that the distance from a given transmit-receive antenna pair may exhibit as multiple candidates. We may formulate the likelihood as a mixture of Gaussian (MoG), where each Gaussian is centered at a different integer wavelength. In particular, given the true distance d, we may approximate the distance estimates d est with the following:
- N is determined by the total number of candidates
- i is integer wavelength
- ⁇ w corresponds to the standard deviation of Gaussian noise.
- weights w(n) denote the discrete distribution of different integers. For instance, these weights may correspond to the value of the fractional Fourier transform or a MUSIC projection. In some cases: (a) the overall algorithm performance is the same in both cases; and (b) the fractional Fourier is employed because it is less computationally expensive.
- the LS employs multiple receivers, each of which provides a different set of distance estimates.
- estimates from different antennas may be independent, and the overall likelihood p(y t
- Equations 2 and 3 may form the HMM.
- the LS performs approximate inference solutions, such as particle filters. Instead of propagating beliefs through analytical probability distributions, the LS may employ models that approximate distributions with a set of particles.
- intersection points between ellipses are from different antennas. However, this is undesirable for three reasons. First, any such two-way intersection point leaves out important information about distance measurements from other receive antennas. Second, it would result in a polynomially large number of intersection points. Third, such a set of intersection points would not be representative of the distribution as they are sparsely distributed rather than concentrated at the most likely locations.
- the LS identifies the most likely intersection points and then uses them to sample particles that are representative of the distribution.
- the LS may exploit the fact that each antenna's distance measurements may be modeled as a Mixture of Gaussians in order (1) to prune out unlikely candidates, and (2) to linearize the intersection points into a mixture of 2D Gaussians. Subsequently, the LS may sample from a mixture of 2D Gaussians and then adapt the particle weights by importance sampling.
- Step 1 Pruning.
- the LS may use the distances between intersection points to identify the most likely candidates. For instance, given a tuple of distance measurements from three antennas, [d 0 , d 1 , d 2 ], the LS may calculate three ellipses and identify the three closest intersection points between them as shown in FIG. 8 . In an ideal noiseless context, the three closest intersection points would coincide. In a practical, real-world, noisy context, we may expect that the three closest intersection points do not coincide. In some cases, the LS computes the average distance between the three intersection points and discards all intersection points whose average is above a threshold.
- r(s) ⁇ T ⁇ where r ( s ): ⁇ x 01 s ⁇ x 12 s ⁇ 2 + ⁇ x 01 s ⁇ x 02 s ⁇ 2 + ⁇ x 12 s ⁇ x 02 s ⁇ 2 )
- the tuple s is integer tuple [s 0 ,s 1 ,s 2 ] for [d 0 +s 0 ⁇ ,d 1 +s 1 ⁇ ,d 2 +s 2 ⁇ ]).
- Step 2 Geometric Linearization.
- the LS approximates each intersection point as a 2D Gaussian using the original distance distributions of the corresponding ellipses.
- this linearization is possible, consider two intersecting ellipses in FIG. 9 , whose intersection point we denote as x 01 s . Since the noise on the distance measurements is much smaller than the size of the ellipses, the curvature of the ellipses around the intersection point is relatively flat. This allows the LS to approximate the distribution as a 2D Gaussian centered around x 01 s and whose axes lie along the normals to the two ellipses at the intersection point.
- the LS may obtain R ij as a Gaussian distribution in the 2D coordinate system:
- FIG. 9 shows normals at an intersection of ellipses.
- ellipses 901 and 902 intersect at intersection 903 .
- Normals a and b are normals to ellipses 902 and 901 , respectively, at intersection 903 .
- Normals a and b may be treated as two orthogonal axes of a 2D space.
- the curvature of the ellipses is relatively flat (compared to the noise on distance measurements).
- the distribution of probabilities of candidate location may be modeled as a 2D Gaussian that is centered around intersection 903 and that has axes that lie along normals a and b to the two ellipses 901 , 902 at intersection 903 .
- Step 3 Centroid Approximation. So far, the LS may have identified the most likely tuples of intersection points, and linearized each of the intersection points as a 2D Gaussian. Next, in some cases, the LS fuses every surviving tuple into a single candidate. This step is helpful because it may combine measurements from all antennas (while the previous step considered antennas only in pairs). To do so, the LS may approximate any given candidate x 0 as the centroid of its three corresponding intersection points (from Step 1).
- This provides us with a set of 2D Gaussian distributions, each centered at a likely candidate position.
- FIG. 8 illustrates centroid approximation.
- three ellipses 821 , 822 , 823 intersect at a total of six intersection points, including a first subset of intersection points 801 , 802 , 803 .
- This first subset is the three intersection points which are closest to each other. Specifically, the average pairwise distance between intersections in this first subset is less than the average pairwise distance between intersections for any other subset of three intersections (in the set of six intersections).
- a centroid 804 of the first subset of intersection points 801 , 802 , 803 may be computed.
- Step 4 Weighting and Initial Sampling.
- the LS may combine the set of Gaussian distributions into a single 2D mixture of Gaussians. To do so, the LS may assign weights proportional to the product of weights of individual integers to tuples in C. This provides an initial sampling distribution q 0 (x 0 ;y 0 ):
- the LS may efficiently sample a small number of particles ⁇ x 0 (i) ⁇ and use them to approximate the distribution. After sampling, the LS may normalize every particle's weight to the total weights.
- FIGS. 10A and 10B illustrate a particle filter converging on an accurate estimate of location.
- the particles are concentrated in several clusters which correspond to the fused intersection points.
- FIG. 10A shows an initial sampling in 2D space.
- the actual location 1001 is in one of the clusters, but the particle with the highest weight 1002 belongs to another cluster. This shows the power of fusing across antennas but also emphasizes the need to fuse measurements over time.
- the highest weight particle may converge to the correct one.
- the actual location 1001 and the particle with the highest weight 1002 have converged very close to each other and are located in the same cluster.
- the LS may update its particles through new observations over time.
- the LS may employ Sequential Importance Sampling (SIS), a type of particle update filter.
- SIS Sequential Importance Sampling
- the SIS may operate through a sequential process.
- the SIS may use the (new) observations from time t and the probability distribution from the previous state t ⁇ 1.
- it may be optimal to sample from the following distribution p ( x t
- x i-1 (i) ) is a Gaussian as per the motion model in Equation 3.
- x t ) may be approximated as a Gaussian as per Equation 5. Because the product of two Gaussians is Gaussian, the overall sampling function may be both Gaussian and approximately optimal. Note that p(y t
- SIS samples a single particle from that distribution for time t. It may then repeat the same process for all the particles. After obtaining all the samples at time t, it may normalize their weights and then move to the next time step t+1 and repeat the same process.
- the LS may make a decision about the most likely location by choosing the particle with the highest weight. It may also update the past trajectory since every particle may backtrack its entire history. This may allow the LS to continuously update the trajectory with new observations.
- the SIS algorithm may—unless a corrective step is taken—degenerate to a small number of particles.
- the LS may employ a resampling approach. In particular, at every sampled time t, the LS may estimate the effective sample size
- the LS may exploit the fact that many of the particles i come from the same Gaussian in the Gaussian mixture because they share the same tuple of intersection points. Hence, to optimize computation, the LS may compute the sequential sampling function once for each cluster and reuse it for all particles in that cluster.
- Outlier Detection and Recovery In some cases, wireless interference or strong leakage from multipath occurs. To deal with such scenarios, the LS may employ outlier detection and recovery. For instance, outliers may, in some cases, occur: (a) where the centroid at t is far from that at t ⁇ 1; or (b) where all particles are assigned near-zero weights, indicating that there is no valid position given current distance estimation. To recover from such outliers, the LS may leverage its high frame rate and interpolate the previous estimates.
- FIGS. 11A and 11B together comprise a flowchart of a localization method.
- the method employs an approximate inference algorithm.
- a localization helper is implemented on USRP X310 software radios, all synchronized to the same external clock.
- Three USRPs are employed with two UBX daughterboards each.
- Each USRP may support two chains. In the prototype, one chain is used as a transmitter and four as receivers. Each transmit/receive chain is connected to a circularly polarized patch antenna. All antennas are arranged in a single plane, with a separation of 40 cm between any two adjacent pairs.
- USRPs sample at 100 MSps and send data over Ethernet to a computer using the Intel® Converged Network Adapter X520-DA2 to support high data rates.
- the computer runs Ubuntu® 16.04 and has a 4-core 8-thread 64-bit Intel® Core i7 processor and 16 GB RAM.
- a USRP RFID reader is implemented on USRP N210 software radios with SBX daughterboards.
- the reader implements the EPC-Gen 2 RFID protocol.
- the reader powers up and communicate with the RFID tags, setting various parameters such as the BLF and tag selection in multi-tag scenarios.
- the reader employs commands so that it only needs to decode the RN16 of the RFIDs of interest; this enables it to skip the ACK and EPC response portions of the communication session and achieve higher frame rates.
- the LS performs real-time processing.
- the localization algorithm is encoded in the USRP's UHD driver in C++.
- a multi-threaded pipelined architecture with worker pools is employed.
- On the transmit side two threads in total are employed: one for the RFID reader and another one for the OFDM transmitter.
- On the receive side one thread is employed for extracting buffer-sized packets from the USRPs.
- every antenna has a worker pool of two threads; each worker extracts and processes individual RFID responses from the shared buffer, computes and timestamps the wideband estimates, and feeds them to an output buffer.
- the threads are aggregated, combining the wideband estimates from all the antennas to perform 3D localization.
- the prototype tracked the location of commercial, off-the-shelf, battery-free UHF RFID tags.
- the prototype was evaluated with multiple types of RFID tags. Because the tags were vertically polarized, the robots were programmed to maintain their orientation to minimize phase changes from orientation changes. Alternatively, circularly polarized tags may be employed to avoid this problem, or an orientation-phase model may be incorporated into the LS's inference algorithm.
- the prototype detected position of RFIDs on three kinds of moving robots.
- an RFID was attached to a cleaning Roomba® robot.
- an RFID was attached to an item carried by LeArm 6DoF robotic arms and to Potensic® A20 minidrones.
- the prototype was evaluated in a standard office building, fully furnished with tables, chairs, computers.
- the prototype was tested in both line-of-sight (LOS) and non-line-of-sight (NLOS) settings.
- the NLOS testing was performed by blocking the visible LOS path between an RFID tag and the prototype's antenna using standard office cubicle dividers made of wood.
- the prototype's localization accuracy was tested in LOS and NLOS settings. Five thousand experimental trials were performed, each lasting between 30 seconds and one minute. This allowed more than 20 million location measurements to be collected. In each trial, different arbitrary movements were performed with the robotic arm or the drone carrying the RFID. The tracking error was computed as the difference between the prototype's location estimate and ground truth as determined by an OptiTrack® optical tracking system.
- the prototype achieved a median error of less than 1 cm and a 90th percentile error of less than 2 cm in each dimension.
- the latency of the prototype was evaluated for both 2D and 3D localization.
- Our 2D trials were performed using an RFID attached to a Roomba® robot.
- the latency was computed as the time difference between the time the USRP obtains an RFID response and the time it outputs a location.
- the time-of-flight of the wireless signal was ignored (for purposes of evaluating test results) since it is only few nanoseconds for the distances of testing (few meters) hence negligible for our latency measurements.
- the test results show that the 99th percentile latency measurements for 2D and 3D tracking by the prototype are 6.6 cms and 7.3 ms, respectively. This latency is primarily limited by the processing time of the computer rather than the latency of the RFID's response or acquisition.
- the prototype's receivers all obtain the RFID responses at the same time, yet the difference in latency may be due to the difference in processing speed.
- the prototype achieved a frame rate of 300 frames/second in both 2D and 3D localization. This high frame rate may owe to the prototype's pipelined architecture which decouples the frame rate from the latency, enabling it to achieve both high frame rates and low latency.
- the prototype's high accuracy, low latency, and high frame rate enable it to track objects in continuous motion. This ability to track an object in continuous motion is helpful for robotic manipulation and tracking tasks.
- the prototype's accuracy with motion was evaluated. Four hundred experimental trials were performed, each time varying the speed at which an RFID tag moves. The speed was varied by attaching the RFID tag to a mobile Roomba® robot whose speed may be controlled. Even at speeds of 50 cm/s, which is the maximum speed of the Roomba® robot employed in this test, the prototype's error remains under 2 centimeters. This may be due to three reasons: First, the prototype has low latency and high frame rates. Second, the prototype is resilient to multipath due to its larger bandwidth. And lastly, the prototype's Bayesian framework boosts its accuracy and robustness.
- the prototype described in the preceding 12 paragraphs is a non-limiting example of this invention.
- This invention may be implemented in many different ways, and the performance (e.g., accuracy for a static object, latency, and accuracy with motion) of the invention may vary depending on, among other things, the particular implementation of this invention and the conditions in which the invention is employed.
- This invention is not limited to detecting position of battery-free RFID tags that are wirelessly powered by an RFID reader.
- This invention may be employed to detect the position or trajectory of any backscatter node or of any other source of RF radiation.
- this invention may be employed to detect the location of battery-assisted or solar-powered RFID tags.
- Battery-assisted or solar-powered RFID tags may enable communication over relatively long distances. For instance, such tags may be attached to nanodrones or robotic arms to track the robots themselves over tens to hundreds of meters and not just the items they manipulate.
- the wideband signal transmitted or received by the wideband transceiver is an OFDM signal that includes multiple subcarriers.
- the wideband transceiver is sometimes called a localization helper herein.
- the wideband signal is a spread spectrum signal, such as an CDMA (code-division multiple access) signal, SSMA (spread spectrum multiple access) signal or a DSSS (direct-sequence spread spectrum) signal.
- the LS may employ a large bandwidth to obtain “sanitized phases” by projecting on the direct path after identifying it.
- this invention comprises one or more computer that are programmed to perform one or more of the Computer Tasks (as defined herein), including performing any localization algorithm described herein.
- this invention comprises one or more machine readable media, with instructions encoded thereon for one or more computers to perform one or more of the Computer Tasks, including performing any localization algorithm described herein.
- this invention comprises participating in a download of software, where the software comprises instructions for one or more computers to perform one or more of the Computer Tasks, including performing any localization algorithm described herein.
- the participating may comprise (a) a computer providing the software during the download, or (b) a computer receiving the software during the download.
- one or more computers are programmed or specially adapted to perform one or more of the following tasks: (1) to control the operation of, or interface with, hardware components of a localization system, including any RFID reader, wideband transceiver, or localization helper; (2) to estimate channels; (3) to perform a localization algorithm; (4) to receive data from, control, or interface with one or more sensors; (5) to perform any other calculation, computation, program, algorithm, or computer function described or implied herein; (6) to receive signals indicative of human input; (7) to output signals for controlling transducers for outputting information in human perceivable format; (8) to process data, to perform computations, and to execute any algorithm or software; and (9) to control the read or write of data to and from memory devices (tasks 1-9 of this sentence referred to herein
- the one or more computers may, in some cases, communicate with each other or with other devices: (a) wirelessly, (b) by wired connection, (c) by fiber-optic link, or (d) by a combination of wired, wireless or fiber optic links.
- the one or more computers may include, for instance, controller 203 and computer 220 and may also include computers that are housed in, or that are components of, RFID reader 201 , wideband transceiver 202 , wideband transceiver 301 , transmitter 401 or receiver 403 .
- one or more computers are programmed to perform any and all calculations, computations, programs, algorithms, computer functions and computer tasks described or implied herein.
- a machine-accessible medium has instructions encoded thereon that specify steps in a software program; and (b) the computer accesses the instructions encoded on the machine-accessible medium, in order to determine steps to execute in the program.
- the machine-accessible medium may comprise a tangible non-transitory medium.
- the machine-accessible medium comprises (a) a memory unit or (b) an auxiliary memory storage device.
- a control unit in a computer fetches the instructions from memory.
- one or more computers execute programs according to instructions encoded in one or more tangible, non-transitory, computer-readable media.
- these instructions comprise instructions for a computer to perform any calculation, computation, program, algorithm, or computer function described or implied herein.
- instructions encoded in a tangible, non-transitory, computer-accessible medium comprise instructions for a computer to perform the Computer Tasks.
- electronic devices e.g., 201 , 202 , 210 , 211 , 212 , 220 , 301 , 401 , 403
- electronic devices are each configured for wireless or wired communication with other devices in a network.
- one or more of these electronic devices each include a wireless module for wireless communication with other devices in a network.
- Each wireless module may include (a) one or more antennas, (b) one or more wireless transceivers, transmitters or receivers, and (c) signal processing circuitry.
- Each wireless module may receive and transmit data in accordance with one or more wireless standards.
- one or more of the following hardware components are used for network communication: a computer bus, a computer port, network connection, network interface device, host adapter, wireless module, wireless card, signal processor, modem, router, cables or wiring.
- one or more computers are programmed for communication over a network.
- one or more computers are programmed for network communication: (a) in accordance with the Internet Protocol Suite, or (b) in accordance with any other industry standard for communication, including any USB standard, ethernet standard (e.g., IEEE 802.3), token ring standard (e.g., IEEE 802.5), wireless standard (including IEEE 802.11 (wi-fi), IEEE 802.15 (bluetooth/zigbee), IEEE 802.16, IEEE 802.20 and including any mobile phone standard, including GSM (global system for mobile communications), UMTS (universal mobile telecommunication system), CDMA (code division multiple access, including IS-95, IS-2000, and WCDMA), or LTE (long term evolution)), or other IEEE communication standard.
- any other industry standard for communication including any USB standard, ethernet standard (e.g., IEEE 802.3), token ring standard (e.g., IEEE 802.5), wireless standard (including IEEE 802.11 (wi-fi), IEEE 802.15 (blue
- Backscatter node means an object that backscatters a radio signal.
- a non-limiting example of a “backscatter node” is an RFID tag that backscatters an RF signal.
- To compute “based on” specified data means to perform a computation that takes the specified data as an input.
- A comprises B, then A includes B and may include other things.
- a digital computer is a non-limiting example of a “computer”.
- An analog computer is a non-limiting example of a “computer”.
- a computer that performs both analog and digital computations is a non-limiting example of a “computer”.
- an event to occur “during” a time period it is not necessary that the event occur throughout the entire time period. For example, an event that occurs during only a portion of a given time period occurs “during” the given time period.
- Non-limiting examples of an “equation”, as that term is used herein, include: (a) an equation that states an equality; (b) an inequation that states an inequality (e.g., that a first item is greater than or less than a second item); (c) a mathematical statement of proportionality or inverse proportionality; and (d) a system of equations.
- a phrase that includes “a first” thing and “a second” thing does not imply an order of the two things (or that there are only two of the things); and (2) such a phrase is simply a way of identifying the two things, respectively, so that they each may be referred to later with specificity (e.g., by referring to “the first” thing and “the second” thing later).
- the equation may (or may not) have more than two terms, and the first term may occur before or after the second term in the equation.
- a phrase that includes a “third” thing, a “fourth” thing and so on shall be construed in like manner.
- a “given” X is simply a way of identifying the X, such that the X may be referred to later with specificity. To say a “given” X does not create any implication regarding X. For example, to say a “given” X does not create any implication that X is a gift, assumption, or known fact.
- X differs from Y “in that” Z means that X differs from Y in at least the aspect of Z.
- Bob differs from Joe “in that” they have different color hair means that Bob and Joe differ in their hair color and may differ in other aspects also.
- a frequency range is “large” means that the frequency range is at least ten megahertz.
- “Largeband signal” means an RF signal that comprises a set of one or more signals, which set of one or more signals (a) consists of a single signal or consists of signals that are simultaneous with each other, and (b) has a frequency range of at least ten megahertz, which range is from (i) the highest 3 dB cutoff frequency of any signal in the set of one or more signals to (ii) the lowest 3 dB cutoff frequency of any signal in the set of one or more signals.
- frequency range does not mean a range of frequencies through which a signal may vary over time.
- LS Localization system
- LS means a system that is configured to detect 1D, 2D, or 3D position of an object.
- a or B is true if A is true, or B is true, or both A and B are true.
- a calculation of A or B means a calculation of A, or a calculation of B, or a calculation of A and B.
- a parenthesis is simply to make text easier to read, by indicating a grouping of words.
- a parenthesis does not mean that the parenthetical material is optional or may be ignored.
- a “portion” of X means part or all of X.
- Radio band means the frequency band from 3 hertz to 3 terahertz.
- RFID radio frequency identification
- RF transmitter means a transmitter configured to transmit a wireless RF signal.
- a non-limiting example of an “RF transmitter” is a transceiver that is configured to transmit a wireless RF signal.
- RF receiver means a receiver configured to receive a wireless RF signal.
- a non-limiting example of an “RF receiver” is a transceiver that is configured to receive a wireless RF signal.
- RF signal means a signal that has a global maximum of spectral density in the radio band.
- the term “set” does not include a group with no elements.
- Non-limiting examples of a “source” of RF signals are: (a) a RF transmitter; (b) a reflector that is configured to reflect RF signals; and (c) a backscatter node that is configured to backscatter RF signals.
- a “subset” of a set consists of less than all of the elements of the set.
- Transitioning between” a first state and a second state means transitioning from the first state to the second state or transitioning from the second state to the first state.
- a machine-readable medium is “transitory” means that the medium is a transitory signal, such as an electromagnetic wave.
- 3D means three-dimensional.
- the method includes variations in which: (1) steps in the method occur in any order or sequence, including any order or sequence different than that described herein; (2) any step or steps in the method occur more than once; (3) any two steps occur the same number of times or a different number of times during the method; (4) any combination of steps in the method is done in parallel or serially; (5) any step in the method is performed iteratively; (6) a given step in the method is applied to the same thing each time that the given step occurs or is applied to different things each time that the given step occurs; (7) one or more steps occur simultaneously, or (8) the method includes other steps, in addition to the steps described herein.
- any definition or clarification herein of a term or phrase applies to any grammatical variation of the term or phrase, taking into account the difference in grammatical form.
- the grammatical variations include noun, verb, participle, adjective, and possessive forms, and different declensions, and different tenses.
- this invention is a system comprising: (a) one or more radio frequency (RF) transmitters; (b) one or more RF receivers; and (c) one or more computers; wherein (i) the one or more transmitters are configured to wirelessly transmit a largeband signal in such a way that the largeband signal travels from the one or more transmitters to a backscatter node, reflects from the backscatter node and travels to the one or more RF receivers, (ii) the one or more RF receivers are each configured to take measurements of the largeband signal, (iii) the one or more computers are programmed (A) to identify, based on the measurements, when the backscatter node is in a first reflective state and when the backscatter node is in a second reflective state, the first reflective state differing from the second reflective state in that a reflection response of the backscatter node is different in the first reflective state than in the second reflective state, which reflection response is phase or amplitude of RF reflection
- the one or more computers are programmed to determine a one-dimensional, two-dimensional or three-dimensional position of the backscatter node, based on the estimates of phase or amplitude.
- the system further includes an additional transmitter, which additional transmitter is configured to transmit an additional signal; and (b) the system is configured to transmit the largeband signal in such a way that the largeband signal reflects from the backscatter node while the reflection response of the backscatter node is modulated by the backscatter node in response to the additional signal.
- the one or more computers are programmed to cause the one or more transmitters to transmit the largeband signal in such a way that transmitting the largeband signal comprises orthogonal frequency-division multiplexing.
- the one or more computers are programmed: (a) to cause the one or more transmitters to transmit the largeband signal in such a way that transmitting the largeband signal comprises orthogonal frequency-division multiplexing (OFDM); and (b) to compute, for each specific frequency band in the set of frequency bands, an average phase and average amplitude, by subtracting a first average phase from a second average phase and by subtracting a first average amplitude from a second average amplitude, where (i) the first average phase and first average amplitude are average phase and average amplitude, respectively, for the specific frequency band as measured by the one or more receivers during OFDM symbols that occur only while the backscatter node is in a non-reflective state, and (ii) the second average phase and second average amplitude are average phase and average amplitude, respectively, for the specific frequency band as measured by the one or more receivers during OFDM symbols that occur only while the backscatter node is in a reflective state.
- OFDM orthogonal frequency-d
- the system further comprises a radio frequency identification (RFID) reader; and (b) the backscatter node is an RFID tag.
- the one or more RF transmitters transmit a set of orthogonal frequency-division multiplexing (OFDM) symbols in such a way that: (a) each OFDM symbol has a duration; and (b) the duration is sufficiently short that each OFDM symbol in a subset of the set of OFDM symbols (i) occurs only while the backscatter node remains in the first reflective state or remains in the second reflective state, and (ii) does not occur while the backscatter node is transitioning between the first and second reflective states.
- OFDM orthogonal frequency-division multiplexing
- the system further comprises a radio frequency identification (RFID) reader;
- the backscatter node is an RFID tag;
- the RFID tag is configured to transmit a preamble to an RFID communication;
- the one or more computers are programmed (i) to compute, for each specific time segment in a set of time segments, a correlation between (A) a subset of the measurements, which subset was taken during the specific time segment, and (B) data that encodes the preamble and that was stored in memory, (ii) to identify, as a calculated window for the preamble, a time segment for which the correlation is at least as great as that for any other time segment in the set of time segments, (iii) to determine, based on a switching rate of the RFID tag and on timing of the calculated window for the preamble, a set of transition time intervals in which the RFID tag is transitioning between the first and second reflective states, and (iv) to disregard the set of transition time intervals, when estimating the position.
- RFID radio frequency identification
- this invention is a system comprising: (a) one or more radio frequency (RF) receivers that are configured to receive a largeband signal; and (b) one or more computers; wherein the one or more computers are programmed to perform an estimation, which estimation includes (i) calculating estimates of phase or amplitude of the largeband signal for each frequency band in a set of multiple frequency bands of the largeband signal, which estimates are based on measurements (A) taken by the one or more RF receivers during a set of multiple time periods in such a way that measurements for each time period in the set of time periods are separate from measurements for each other time period in the set of time periods, or (B) taken at a set of multiple receiver antennas of the one or more RF receivers in such a way that measurements taken at each antenna in the set of antennas are separate from measurements taken at each other antenna, if any, in the set of antennas, (ii) identifying, based on bandwidth of the largeband signal, a maximum distance and a minimum distance, (iii)
- the estimation includes performing Bayesian inference.
- the one or more computers are further programmed to perform modeling, which modeling: (a) is based on the set of estimates of phase or amplitude; and (b) models a probability distribution of the candidate distances as a mixture of gaussians, where each gaussian is centered at a different integer multiple of a wavelength.
- the estimation includes solving a Hidden Markov Model.
- at least a subset of the measurements are gaussian related to each other over time.
- the estimation includes solving a maximum a posteriori problem to estimate a Hidden Markov Model.
- the estimation involves solving a maximum a posteriori problem by performing approximate inference, which approximate inference includes a particle filter.
- the estimation includes: (a) identifying (i) a set of ellipses that includes, for each specific receiver antenna in the set of receiver antennas, ellipses that each have foci, which foci are the specific receiver antenna and a transmitter antenna, and (ii) a set of most likely points of intersection between ellipses in the set of ellipses; or (b) identifying (i) a set of ellipsoids that includes, for each specific receiver antenna in the set of receiver antennas, ellipsoids that each have foci, which foci are the specific receiver antenna and a transmitter antenna, and (ii) a set of most likely points of intersection between ellipsoids in the set of ellipsoids.
- the estimation includes: (a) identifying a set of ellipses that includes, for each specific receiver antenna in the set of receiver antennas, ellipses that each have foci, which foci are the specific receiver antenna and a transmitter antenna; (b) identifying a set of most likely points of intersection between ellipses in the set of ellipses; and (c) estimating, based on the measurements, a first subset of n intersection points in the set of most likely intersection points, where the sum of pairwise distances between intersection points in the first subset is less than or equal to the sum of pairwise distances between intersection points in any other subset of n intersection points in the set of most likely intersection points.
- the estimation involves performing an approximate inference that comprises a particle filter; and (b) the particle filter includes updating particles by sequential importance sampling, to update for new measurements of the largeband signal.
- the one or more computers are further programmed to replace each outlier estimate, in a set of outlier estimates of position, with an interpolated estimate of position, which interpolated estimate is interpolated from a previous position.
- this invention is a system comprising: (a) one or more radio frequency (RF) transmitters; (b) one or more RF receivers; and (c) one or more computers; wherein (i) the one or more transmitters are configured to wirelessly transmit a largeband signal in such a way that the largeband signal travels from the one or more transmitters to a backscatter node, reflects from the backscatter node and travels to the one or more RF receivers, (ii) the system is configured to take measurements of the largeband signal at each of multiple receiver antennas of the one or more RF receivers, (iii) the one or more computers are programmed (A) to identify, based on the measurements, when the backscatter node is in a first reflective state and when the backscatter node is in a second reflective state, the first reflective state differing from the second reflective state in that a reflection response of the backscatter node is different in the first reflective state than in the second reflective state, which reflection response is
- each signal described above may comprise a largeband signal.
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Discrete Mathematics (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Radar Systems Or Details Thereof (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
Description
For example, if the helper's OFDM bandwidth is 100 MHz and the backscatter link frequency is 500 KHz, the LS may employ OFDM where N is less than 100.
Robust Wideband Channel Estimation
where T is the preamble length and A is the time instance where correlation is performed.
resolution=speed/bandwidth
where d is the distance traveled and λ is the wavelength.
where xt is a d dimensional coordinate vector of the tag, yt is a k dimensional vector of the estimated distances from each receiver, and p(x|y) denotes the posterior probability distribution.
where N is determined by the total number of candidates, i is integer wavelength, and σw corresponds to the standard deviation of Gaussian noise.
where dist: Rd→Rk returns the round trip distance to the RFID, k denotes the total number of receivers, and j is the index of each receiver.
Transition Model.
x t =x t-1 +v t ,v t˜(0,Σv) (Eq. 3)
where vt 's are i.i.d. Gaussian sequences, accounting for the position changes, and Σv denotes their standard deviation.
r(s):=∥x 01 s −x 12 s∥2 +∥x 01 s −x 02 s∥2 +∥x 12 s −x 02 s∥2)
where V denotes the diagonal covariance matrix, which represents the fact that the distance errors from the different antennas are independent.
f(x 0 ;y 0 ,s)˜(μs,Σs)
μs=⅓(x 01 s +x 12 s +x 02 s)
Σs=(⅓)2(Σ01 s+Σ12 s+Σ02 s) (Eq. 5)
p(x t |y t-1 (i) ,y i)∝p(y t |x t)p(x t |x t-1 (i))
and may resample when the effective sample size is lower than a threshold.
Claims (15)
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US16/162,085 US10575277B2 (en) | 2017-11-08 | 2018-10-16 | Methods and apparatus for wideband localization |
US16/727,727 US10701662B2 (en) | 2017-11-08 | 2019-12-26 | Methods and apparatus for wideband localization |
US16/874,684 US20200280953A1 (en) | 2017-11-08 | 2020-05-15 | Methods and Apparatus for Wideband Localization |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US201762582949P | 2017-11-08 | 2017-11-08 | |
US16/162,085 US10575277B2 (en) | 2017-11-08 | 2018-10-16 | Methods and apparatus for wideband localization |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/727,727 Continuation US10701662B2 (en) | 2017-11-08 | 2019-12-26 | Methods and apparatus for wideband localization |
Publications (2)
Publication Number | Publication Date |
---|---|
US20190141666A1 US20190141666A1 (en) | 2019-05-09 |
US10575277B2 true US10575277B2 (en) | 2020-02-25 |
Family
ID=66327972
Family Applications (3)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/162,085 Active US10575277B2 (en) | 2017-11-08 | 2018-10-16 | Methods and apparatus for wideband localization |
US16/727,727 Active US10701662B2 (en) | 2017-11-08 | 2019-12-26 | Methods and apparatus for wideband localization |
US16/874,684 Abandoned US20200280953A1 (en) | 2017-11-08 | 2020-05-15 | Methods and Apparatus for Wideband Localization |
Family Applications After (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/727,727 Active US10701662B2 (en) | 2017-11-08 | 2019-12-26 | Methods and apparatus for wideband localization |
US16/874,684 Abandoned US20200280953A1 (en) | 2017-11-08 | 2020-05-15 | Methods and Apparatus for Wideband Localization |
Country Status (1)
Country | Link |
---|---|
US (3) | US10575277B2 (en) |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10575277B2 (en) * | 2017-11-08 | 2020-02-25 | Massachusetts Institute Of Technology | Methods and apparatus for wideband localization |
US11763106B2 (en) | 2018-06-05 | 2023-09-19 | The Research Foundation For The State University Of New York | Method for passive wireless channel estimation in radio frequency network and apparatus for same |
US11035947B2 (en) * | 2018-11-06 | 2021-06-15 | Aerial Surveying, Inc | Radio frequency (RF) ranging in propagation limited RF environments |
CN109917909B (en) | 2019-02-01 | 2022-05-31 | 成都思悟革科技有限公司 | Motion capture device and method of multi-point receiving array based on non-propagation electromagnetic field |
US11386279B2 (en) * | 2019-04-19 | 2022-07-12 | Nec Corporation | Efficient decoding of RFID tags based on relevant RN16 selection |
CN112532342B (en) * | 2019-09-17 | 2023-05-16 | 华为技术有限公司 | Data transmission method and device in back reflection communication |
US11290960B2 (en) | 2020-02-10 | 2022-03-29 | Huawei Technologies Co., Ltd. | Method and apparatus for low power transmission using backscattering |
CN111654314B (en) * | 2020-06-01 | 2022-06-24 | 电子科技大学 | Multi-reflection equipment symbiotic wireless communication system |
WO2022101739A1 (en) * | 2020-11-11 | 2022-05-19 | I Comb Ltd. | Systems and methods for ascertaining spatial information of target objects |
EP4289192A1 (en) | 2021-03-22 | 2023-12-13 | Nokia Technologies Oy | Use of sidelink communications for backscatter node positioning within wireless networks |
CN113438133B (en) * | 2021-06-30 | 2022-11-29 | 北京小米移动软件有限公司 | UWB device test method, device, system, electronic device and storage medium |
CN113453336B (en) * | 2021-08-31 | 2022-01-25 | 广东省新一代通信与网络创新研究院 | Communication positioning method, base station and terminal equipment |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030052813A1 (en) * | 2001-09-17 | 2003-03-20 | Kazuma Natsume | Radar designed to minimize error in detecting target |
US6750960B2 (en) * | 2001-03-06 | 2004-06-15 | The Regents Of The University Of California | Optical distance measurement device and method thereof |
US20060261951A1 (en) | 2005-04-26 | 2006-11-23 | Rf Code, Inc. | RFID systems and methods employing infrared localization |
US20070115137A1 (en) | 2005-11-21 | 2007-05-24 | Lyon Geoff M | Method and apparatus for localization of RFID tags |
US20070285245A1 (en) | 2006-05-01 | 2007-12-13 | Djuric Petar M | RFID system and method for localizing and tracking a moving object with an RFID tag |
US20090212921A1 (en) | 2008-02-25 | 2009-08-27 | Wirama Corporation | Localizing tagged assets using modulated backscatter |
US20100309017A1 (en) | 2009-03-11 | 2010-12-09 | Checkpoint Systems, Inc. | Localization Using Virtual Antenna Arrays In Modulated Backscatter RFID Systems |
US7898419B2 (en) | 2007-07-27 | 2011-03-01 | Lucomm Technologies, Inc. | Systems and methods for object localization and path identification based on RFID sensing |
US20110309969A1 (en) * | 2010-06-17 | 2011-12-22 | Tc License Ltd. | System and method for microwave ranging to a target in presence of clutter and multi-path effects |
US20130285750A1 (en) * | 2012-04-30 | 2013-10-31 | Broadcom Corporation | Supply tracking |
US9043992B2 (en) | 2009-10-20 | 2015-06-02 | Teknologian Tutkimuskeskus Vtt | Method and system for object localization |
US20160097837A1 (en) * | 2014-06-05 | 2016-04-07 | Zih Corp. | Method for Iterative Target Location in a Multiple Receiver Target Location System |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10575277B2 (en) * | 2017-11-08 | 2020-02-25 | Massachusetts Institute Of Technology | Methods and apparatus for wideband localization |
-
2018
- 2018-10-16 US US16/162,085 patent/US10575277B2/en active Active
-
2019
- 2019-12-26 US US16/727,727 patent/US10701662B2/en active Active
-
2020
- 2020-05-15 US US16/874,684 patent/US20200280953A1/en not_active Abandoned
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6750960B2 (en) * | 2001-03-06 | 2004-06-15 | The Regents Of The University Of California | Optical distance measurement device and method thereof |
US20030052813A1 (en) * | 2001-09-17 | 2003-03-20 | Kazuma Natsume | Radar designed to minimize error in detecting target |
US20060261951A1 (en) | 2005-04-26 | 2006-11-23 | Rf Code, Inc. | RFID systems and methods employing infrared localization |
US20070115137A1 (en) | 2005-11-21 | 2007-05-24 | Lyon Geoff M | Method and apparatus for localization of RFID tags |
US20070285245A1 (en) | 2006-05-01 | 2007-12-13 | Djuric Petar M | RFID system and method for localizing and tracking a moving object with an RFID tag |
US7898419B2 (en) | 2007-07-27 | 2011-03-01 | Lucomm Technologies, Inc. | Systems and methods for object localization and path identification based on RFID sensing |
US20090212921A1 (en) | 2008-02-25 | 2009-08-27 | Wirama Corporation | Localizing tagged assets using modulated backscatter |
US20100309017A1 (en) | 2009-03-11 | 2010-12-09 | Checkpoint Systems, Inc. | Localization Using Virtual Antenna Arrays In Modulated Backscatter RFID Systems |
US9043992B2 (en) | 2009-10-20 | 2015-06-02 | Teknologian Tutkimuskeskus Vtt | Method and system for object localization |
US20110309969A1 (en) * | 2010-06-17 | 2011-12-22 | Tc License Ltd. | System and method for microwave ranging to a target in presence of clutter and multi-path effects |
US20130285750A1 (en) * | 2012-04-30 | 2013-10-31 | Broadcom Corporation | Supply tracking |
US20160097837A1 (en) * | 2014-06-05 | 2016-04-07 | Zih Corp. | Method for Iterative Target Location in a Multiple Receiver Target Location System |
Non-Patent Citations (19)
Title |
---|
Azzouzi, S., et al., New measurement results for the localization of UHF RFID transponders using an Angle of Arrival (AoA) approach; published in 2011 IEEE International Conference on RFID, 2011. |
Bouet, M. et al., A range-free 3-D localization method for RFID tags based on virtual landmarks; published in IEEE 19th International Symposium on Personal, Indoor and Mobile Radio Communications, 2008. PIMRC 2008. |
Bouet, M., et al., RFID tags: Positioning principles and localization techniques; published in 2008 1st IFIP Wireless Days. |
Chawla, K., et al., Real-time RFID localization using RSS; published in 2013 International Conference on Localization and GNSS (ICL-GNSS), 2013. |
Denk, W., et al., Two-photon laser scanning fluorescence microscopy; published in Science, vol. 248, No. 4951, pp. 73-76, Apr. 1990. |
Helmchen, F., et al., Deep tissue two-photon microscopy; published in Nature Methods, vol. 2, No. 12, pp. 932-940, Dec. 20005. |
Karmakar, N., et al., Chipless RFID Tag Localization; published in IEEE Transactions on Microwave Theory and Techniques, vol. 61 Issue 11, Nov. 2013. |
Kronberger, R., et al., UHF RFID localization system based on a phased array antenna; published in 2011 IEEE International Symposium on Antennas and Propagation (APSURSI), 2011, pp. 525-528. |
Ma, Y. et al., 3D real-time indoor localization via broadband nonlinear backscatter in passive devices with centimeter precision; published in Proceedings of the 22nd Annual International Conference on Mobile Computing and Networking MobiCom '16, pp. 216-229, Oct. 2016, ACM. |
Ni, L., et al. LANDMARC: indoor location sensing using active RFID; published in Proceedings of the First IEEE International Conference on Pervasive Computing and Communications, 2003. (PerCom 2003). |
Pelissier, M., et al., A 112 Mb/s Full Duplex Remotely-Powered Impulse-UWB RFID Transceiver for Wireless NV-Memory Applications; published in IEEE Journal of Solid-State Circuits; Year 2011, vol. 46, Issue 4, pp. 916-927. |
Shangguan, L., et al., The Design and Implementation of a Mobile RFID Tag Sorting Robot; published in MobiSys 16, Proceedings of the 14th Annual International Conference on Mobile Systems, Applications, and Services pp. 31-42. |
Yang, L., et al., Tagoram: real-time tracking of mobile RFID tags to high precision using COTS devices; published in MobiCom '14 Proceedings of the 20th annual international conference on Mobile computing and networking pp. 237-248, Sep. 2014. |
Yang, Z., et al., NICScatter: Backscatter as a Covert Channel in Mobile Devices; published in MobiCom '17 Proceedings of the 23rd Annual International Conference on Mobile Computing and Networking, pp. 356-367,-Oct. 2017. |
Yang, Z., et al., NICScatter: Backscatter as a Covert Channel in Mobile Devices; published in MobiCom '17 Proceedings of the 23rd Annual International Conference on Mobile Computing and Networking, pp. 356-367,—Oct. 2017. |
Zhou, C. et al., Accurate Phase-Based Ranging Measurements for Backscatter RFID Tags; published in IEEE Antennas and Wireless Propagation Letters; Year 2012, vol. 11, pp. 152-155. |
Zhou, J. et al., RFID localization algorithms and applications-a review; published in Journal of Intelligent Manufacturing; vol. 20, Issue 6, (Dec. 2009), pp. 695-707. |
Zhou, J. et al., RFID localization algorithms and applications—a review; published in Journal of Intelligent Manufacturing; vol. 20, Issue 6, (Dec. 2009), pp. 695-707. |
Zhou, J., et al., Two-dimension localization of passive RFID tags using AOA estimation; published in 2011 IEEE Instrumentation and Measurement Technology Conference (I2MTC), 2011. |
Also Published As
Publication number | Publication date |
---|---|
US10701662B2 (en) | 2020-06-30 |
US20200280953A1 (en) | 2020-09-03 |
US20190141666A1 (en) | 2019-05-09 |
US20200145959A1 (en) | 2020-05-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10701662B2 (en) | Methods and apparatus for wideband localization | |
Luo et al. | 3d backscatter localization for {Fine-Grained} robotics | |
Großwindhager et al. | SALMA: UWB-based single-anchor localization system using multipath assistance | |
Karanam et al. | Magnitude-based angle-of-arrival estimation, localization, and target tracking | |
Xie et al. | mD-Track: Leveraging multi-dimensionality for passive indoor Wi-Fi tracking | |
Qian et al. | Widar2. 0: Passive human tracking with a single Wi-Fi link | |
Yang et al. | See through walls with COTS RFID system! | |
US11604250B2 (en) | Methods and apparatus for wideband localization | |
US20170212210A1 (en) | Wireless positioning systems | |
Qian et al. | Decimeter level passive tracking with WiFi | |
WO2018160141A1 (en) | Apparatus and method for localisation and/or tracking | |
WO2016011433A2 (en) | Wireless positioning systems | |
HajiAkhondi-Meybodi et al. | Bluetooth low energy and CNN-based angle of arrival localization in presence of Rayleigh fading | |
Soltanaghaei et al. | Tagfi: Locating ultra-low power wifi tags using unmodified wifi infrastructure | |
Storrer et al. | Indoor tracking of multiple individuals with an 802.11 ax Wi-Fi-based multi-antenna passive radar | |
Zou et al. | Joint adversarial domain adaptation for resilient WiFi-enabled device-free gesture recognition | |
Hao et al. | A method for improving UWB indoor positioning | |
An et al. | General-purpose deep tracking platform across protocols for the internet of things | |
Zhu et al. | Calibrating time-variant, device-specific phase noise for COTS WiFi devices | |
Paulino et al. | Self-localization via circular bluetooth 5.1 antenna array receiver | |
Li et al. | Configurable multipath-assisted indoor localization using active relay | |
Ma et al. | Fine-grained RFID localization via ultra-wideband emulation | |
Hoffman et al. | RSS and phase Kalman filter fusion for improved velocity estimation in the presence of real-world factors | |
CA3102603A1 (en) | Method for passive wireless channel estimation in radio frequency network and apparatus for same | |
Zhang et al. | Towards ubiquitous sensing and localization with reconfigurable intelligent surfaces |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
FEPP | Fee payment procedure |
Free format text: ENTITY STATUS SET TO UNDISCOUNTED (ORIGINAL EVENT CODE: BIG.); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
AS | Assignment |
Owner name: MASSACHUSETTS INSTITUTE OF TECHNOLOGY, MASSACHUSET Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:LUO, ZHIHONG;ZHANG, QIPING;SELBY, NICHOLAS;AND OTHERS;SIGNING DATES FROM 20181016 TO 20181017;REEL/FRAME:047760/0471 Owner name: MASSACHUSETTS INSTITUTE OF TECHNOLOGY, MASSACHUSETTS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:LUO, ZHIHONG;ZHANG, QIPING;SELBY, NICHOLAS;AND OTHERS;SIGNING DATES FROM 20181016 TO 20181017;REEL/FRAME:047760/0471 |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NON FINAL ACTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NOTICE OF ALLOWANCE MAILED -- APPLICATION RECEIVED IN OFFICE OF PUBLICATIONS |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: PUBLICATIONS -- ISSUE FEE PAYMENT RECEIVED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: PUBLICATIONS -- ISSUE FEE PAYMENT VERIFIED |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
AS | Assignment |
Owner name: NATIONAL SCIENCE FOUNDATION, VIRGINIA Free format text: CONFIRMATORY LICENSE;ASSIGNOR:MASSACHUSETTS INSTITUTE OF TECHNOLOGY;REEL/FRAME:052252/0946 Effective date: 20181213 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1551); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 4 |